Document not found! Please try again

Development of a Practical Synthesis of an Aminoindanol-Derived M1

Jan 5, 2009 - Air Pollution and Industrial Hygiene · Apparatus and Plant Equipment · Cement, Concrete, and Related ... Marvin M. Hansen*, Sandra S. K...
0 downloads 0 Views 380KB Size
Organic Process Research & Development 2009, 13, 198–208

Development of a Practical Synthesis of an Aminoindanol-Derived M1 Agonist† Marvin M. Hansen,*,‡ Sandra S. K. Borders,‡ Marcella T. Clayton,‡ Perry C. Heath,‡ Stanley P. Kolis,‡ Samuel D. Larsen,‡ Ryan J. Linder,‡ Susan M. Reutzel-Edens,§ Justin C. Smith,‡ Shella L. Tameze,§ Jeffrey A. Ward,‡ and Leland O. Weigel‡ Eli Lilly and Company, Chemical Product Research and DeVelopment DiVision and Pharmaceutical Sciences Research and DeVelopment DiVision, Indianapolis, Indiana 46285, U.S.A.

Abstract: An efficient and scalable synthesis of the clinical candidate 1 is described. The first-generation synthesis built the enantioenriched nitro-aminoindanol core from 6-nitroindanone using a five-step literature route. The second-generation route used a safe aromatic nitration protocol in the presence of the unprotected alcohol to afford the requisite nitro-aminoindanol in one step. Challenges addressed in the remainder of the synthesis include a nitro group reduction to afford ppm levels of unreacted Ar-NO2 (a mutagen) and a novel amidine formation under mild conditions via DMAP/ K2CO3-promoted reaction with a thioimidate-activated amide. A convenient protocol for freebasing the API was provided by stirring with solid K2CO3 and monitoring disappearance of HI by reverse-phase HPLC. Introduction In support of the muscarinic receptor (M1) agonist program at Lilly, we required a practical and scalable synthesis of the aminoindanol-derived clinical candidate 1 (Scheme 1).1 Two

basic routes for assembling the three main subunits were examined. Route A was used to supply the structure-activity relationship (SAR) studies due to facile preparation of a variety of substituted amidines from the Boc-protected aminoindanol 6, followed by deprotection and coupling with a diverse array of biaryl acid chlorides. Aminoindanol 6 was derived from (1R,2R)-trans-1-amino-6-nitro-indan-2-ol (7) via Boc-protection followed by nitro group reduction. Route B was selected for development of a multikilo-scale route in order to avoid the steps required by Boc protection and deprotection. A key component of the synthetic strategy required an improved synthesis of the core aminoindanol subunit 7. Aminoindanol 7 has found importance as a scaffold in several SAR studies2 and has also been utilized as a chiral auxiliary for the asymmetric synthesis of ketones and aldehydes through dynamic kinetic resolution of imines.3 Herein we describe an initial synthesis of candidate 1, utilizing a standard approach to aminoindanol 7, and an improved route that delivered key intermediate 7 in one step from commercially available materials.4

Scheme 1. Retrosynthetic options

† This paper is dedicated to the memory of our friend and former colleague, Dr. Christopher R. Schmid. * Author to whom correspondence may be sent. E-mail: mmh@ lilly.com. ‡ Chemical Product Research and Development Division. § Pharmaceutical Sciences Research and Development Division.

(1) (a) Allen, J. R.; Hitchcock, S. A.; Liu, B.; Turner, W. W.; Jamison, J. A. PCT Int. Appl. WO 2003027061 A2. (b) Bush, J. K.; Heath, P. C. PCT Int. Appl. WO 2004018411 A1. 198 • Vol. 13, No. 2, 2009 / Organic Process Research & Development Published on Web 01/05/2009

(2) (a) Gross, M. F.; Beaudoin, S.; McNaughton-Smith, G.; Amato, G. S.; Castle, N. A.; Huang, C.; Zou, A.; Yu, W. Bioorg. Med. Chem. Lett. 2007, 17, 2849–2853. (b) Kozhushkov, S. I.; Yufit, D. S.; De Meijere, A. AdV. Synth. Catal. 2005, 347, 255–265, references 1-6 cited therein. . (3) Kosmrlj, J.; Weigel, L. O.; Evans, D. A.; Downey, C. W.; Wu, J. J. Am. Chem. Soc. 2003, 125, 3208–3209. (4) Safety note: All nitro aromatic intermediates were found to be mutagenic. Use appropriate care in handling. 10.1021/op800243q CCC: $40.75  2009 American Chemical Society

Scheme 2. Synthesis of 1 via indanone route

Results and Discussion Synthesis of Resolved Aminoindanol 7. The synthesis of the key nitro-aminoindanol 7 from 1-indanone has been reported by several groups,5 and this well precedented route was chosen for an initial kilolab campaign (Scheme 2).6 The nitration of 1-indanone has been reported to provide a 3-4:1 mixture of 6-nitroindanone 8 to the 4-nitroisomer. The isomers have been separated by chromatography2a or by a tedious crystallization in unreported yield. We purchased 50 kg of 6-nitroindanone 8 from Sumikin Chemical Company7 as a 4:1 mixture of 6-nitroand 4-nitro-regioisomers. Since separation at this stage was challenging, the ketone was reduced with sodium borohydride to afford the indanol 9. Crystallization of indanol 9 in the laboratory efficiently rejected the 4-nitro isomer. The process was scaled in the plant in three batches, with the first two batches containing 1.2% of the undesired 4-nitro isomer and the third batch containing 20.1%. Further investigation was required to understand the failure of this crystallization process and ensure utilization of the third pilot-plant lot. Further recrystallization of the 4:1 mixture of 6and 4-nitro isomers from a variety of solvents did not remove, and in some cases enriched, the undesired 4-isomer. Solid-state characterization showed that a new cocrystalline phase was present in the 4:1 mixture relative to the batches with low levels of the 4-nitro isomer. Figure 1 (top) shows two views of the hydrogen-bonding network in the crystal structure of the regiopure, racemic, 6-nitro isomer 9. In Figure 1 (bottom) the hydrogen bonding of the cocrystal obtained by recrystallization of the 4:1 mixture of 6-nitro and 4-nitro isomers (racemic) is (5) (a) Wan, P.; Davis, M.-A.; Teo, J. J. Org. Chem. 1989, 54, 1354– 1359. (b) Buckle, D. R.; Arch, J. R. S.; Edge, C.; Foster, K. A.; HougeFrydrych, C. S. V.; Pinto, I. L.; Smith, D. G.; Taylor, J. F.; Taylor, S. G.; Tedder, J. M.; Webster, R. A. B. J. Med. Chem. 1991, 34, 919– 926. (c) Castle, N. A.; Hollinshead, S. P.; Hughes, P. F.; Mendoza, J. S.; Wilson, J. W. Amato, G.; Beaudoin, S.; Gross, M.; McNaughtonSmith, G. PCT Int. Appl. WO 199904778. (d) See ref 2. (6) Preparation of resolved salt 12 was completed in the pilot plant, and the remaining steps were completed in the kilolab. (7) Sumikin is now Air Water Chemical Co., Ltd.

shown. The formation of a new cocrystal form from the 4:1 mixture and its apparent higher stability (forms in multiple solvent systems) relative to the pure 6-nitro isomer is analogous to the often reported “disappearing polymorphs” phenomenon.8 Having lost our control point for rejection of the 4-nitro isomer, the 4:1 mixture of 9 isomers was evaluated in the downstream chemistry. Elimination of the alcohol with Amberlyst 15 resin in toluene at 100 °C afforded 5-nitroindene (10). The alkene was stable toward isomerization under these conditions in a nonpolar solvent, but other conditions lead to isomerization. NMR solutions in polar solvents such as methanol or DMSO gave isomerization on standing. Addition of Et3N increased the rate of isomerization as expected on the basis of literature studies.9 In this respect, Amberlyst 15 provided a benefit versus the literature protocol using p-toluene sulfonic acid, because aqueous NaHCO3 washes required to remove p-toluene sulfonic acid led to variable amounts of alkene isomerization.2a,5c Nitroindene 10 was not isolated but was carried forward as a toluene/CH2Cl2 solution into an epoxidation with m-CPBA. After a base wash to remove m-CBA, a solvent exchange to heptane resulted in crystallization of epoxide 11. This crystallization efficiently purged the 4-nitro isomer carried in with the alcohol substrate. For the batch with a 4:1 mixture of alcohols, the 4-nitro isomer was controlled to 98.6 wt/wt % assay. Acknowledgment We thank the EML (kilolab) and pilot-plant staff, the analytical R&D group, the pharmaceutical sciences R&D

208



Vol. 13, No. 2, 2009 / Organic Process Research & Development

group, and Robert Waggoner for sourcing support. We thank Tianwei Ma for initial experiments on nitration of cis-aminoindanol, Greg Stephenson for the single-crystal X-ray data, and Julie Bush for final form screening and physical characterization. We thank Stephen Hitchcock and his discovery team for advice and the initial route to prepare amidine 1. Supporting Information Available X-ray crystallographic data for the structures in Figure 1 (6-nitro isomer 9 and the 6-nitro/4-nitro isomeric mixture). This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review September 29, 2008. OP800243Q